Indonesia Conference Directory

Sugarcane quality was evaluated using Near Infrared (NIR) spectroscopy in two sugar factory for payment system. The samples were collected from core sampler system. The parameter of cane quality analysis were brix in cane (BIC), pol in cane (PIC) and fiber in cane. The target of this study was to develop and improve model calibration of NIR for cane quality. NIR reflectance spectra ranging from 1100 – 2500 nm were collected to evaluate the sugarcane quality of shredded cane. The conventional method to determine cane quality was using Wet Disintegrator method based on ICUMSA GS5/7-1 (2011). SX-plus software from Buchi was used as data treatment and model development. The NIR results was evaluate from high correlation coefficient (r2), low standard calibration (SEC) and high ratio of prediction to deviation (RPD). The accuracies of NIR analysis for PIC were satisfactory with coefficient of correlation (R2) above 0,9 both for calibration and prediction. The accuracies of BIC and fiber remained at unsatisfactory level with coefficient of correlation (R2) below 0,9. The NIR instrument could be use to determine pol in cane for cane payment system purposes.